To investigate whether the incorporation of 18FDG-PET into the automatic treatment planning process may be able to decrease the dose to active bone marrow (BM) for locally advanced anal cancer patients undergoing concurrent chemo-radiation (CHT-RT).
Franco et al BMC Cancer (2017) 17:710 DOI 10.1186/s12885-017-3708-4 TECHNICAL ADVANCE Open Access Incorporating 18FDG-PET-defined pelvic active bone marrow in the automatic treatment planning process of anal cancer patients undergoing chemo-radiation Pierfrancesco Franco1*† , Christian Fiandra1†, Francesca Arcadipane1, Elisabetta Trino1, Francesca Romana Giglioli2, Riccardo Ragona1 and Umberto Ricardi1 Abstract Background: To investigate whether the incorporation of 18FDG-PET into the automatic treatment planning process may be able to decrease the dose to active bone marrow (BM) for locally advanced anal cancer patients undergoing concurrent chemo-radiation (CHT-RT) Methods: Ten patients with locally advanced anal cancer were selected Bone marrow within the pelvis was outlined as the whole outer contour of pelvic bones or employing 18FDG-PET to identify active BM within osseous structures Four treatment planning solutions were employed with different automatic optimization approaches toward bone marrow Plan A used iliac crests for optimization as per RTOG 05–29 trial; plan B accounted for all pelvic BM as outlined by the outer surface of external osseous structures; plan C took into account both active and inactive BM as defined using 18FDG-PET; plan D accounted only for the active BM subregions outlined with 18FDG-PET Dose received by active bone marrow within the pelvic (ACTPBM) and in different subregions such as lumbar-sacral (ACTLSBM), iliac (ACTIBM) and lower pelvis (ACTLPBM) bone marrow was analyzed Results: A significant difference was found for ACTPBM in terms of Dmean (p = 0.014) V20 (p = 0.015), V25 (p = 0.030), V30 (p = 0.020), V35 (p = 0.010) between Plan A and other plans With respect to specific subsites, a significant difference was found for ACTLSBM in terms of V30 (p = 0.020)), V35 (p = 0.010), V40 (p = 0.050) between Plan A and other solutions No significant difference was found with respect to the investigated parameters between Plan B,C and D No significant dosimetric differences were found for ACTLSPBM and ACTIBM and inactive BM subregions within the pelvis between any plan solution Conclusions: Accounting for pelvic BM as a whole compared to iliac crests is able to decrease the dose to active bone marrow during the planning process of anal cancer patients treated with intensity-modulated radiotherapy The same degree of reduction may be achieved optimizing on bone marrow either defined using the outer bone contour or through 18FDG-PET imaging The subset of patients with a benefit in terms of dose reduction to active BM through the inclusion of 18FDG-PET in the planning process needs further investigation Keywords: Anal cancer, Hematologic toxicity, Radiotherapy, Dose-painted IMRT, Bone-marrow sparing radiation * Correspondence: pierfrancesco.franco@unito.it † Equal contributors Department of Oncology, Radiation Oncology, University of Turin, Via Genova 3, 10126 Turin, Italy Full list of author information is available at the end of the article © The Author(s) 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated Franco et al BMC Cancer (2017) 17:710 Background At present, concurrent chemo-radiation (CHT-RT) is a standard therapeutic option in patients with squamous cell carcinoma of the anal canal [1, 2] Given the high repopulation rate of this type of tumor, treatment compliance is crucial to avoid unintended interruptions potentially extending overall treatment time [3] In adjunct, maintaining a proper package of chemotherapy (CHT) administration in terms of number of cycles and dose is important to achieve adequate tumor control Hence, decreasing the toxicity profile associated to CHT-RT is crucial If non-conformal techniques are used, as in the RTOG 98–11 trial, crude rates of major acute toxicities can be as high as 48% for skin and 35% for the gastrointestinal district [4] Intensity-modulated radiotherapy (IMRT) provides robust conformality and modulation, abrupt dose fall-off and reliable consistency and may reduce the dose to organs at risk such as bladder, bowel, perineal skin, genitalia and bone marrow, potentially lowering toxicity [5] However, even with this approach, acute toxicity is not negligible, as seen in the RTOG 05–29 trial [6] In this subset of patients, another key endpoint for treatment tolerance is hematologic toxicity (HemT) that can affect compliance to therapy, increasing the likelihood to develop bleeding, infections or fatigue [7] The most important trigger for HemT is CHT that induces myelosuppression [8] Nevertheless, since bone marrow (BM) is highly radiosensitive and, in the average adult population, is comprised for half of its extension within pelvic bones and lumbar vertebrae, the radiation dose received by this compartment may be critical [9, 10] Several retrospective studies correlated different dose parameters of pelvic osseous structures to HT in different oncological scenarios [11–13] Thus, selective sparing of pelvic bones is thought to be a suitable option to decrease HemT during concomitant CHT-RT in patients affected with pelvic malignancies including anal cancer [10] The correct identification of BM within bony structures is the starting point to avoid it during RT Several approaches have been used Contouring the whole bone is the method with the highest chance to be inclusive with respect to BM [11] Delineating the marrow cavity identified as the trabecular bone with lower density on computed tomography is another option [14] The identification of hematopoietically active bone marrow using either magnetic resonance (MR), single-photonemission positron tomography (SPECT), 18F–fluorodeoxyglucose-labeled positron-emission tomography (18FDG-PET) or 3′-deoxy-3′-18F-fluorothymidine-labeled positron-emission tomography (18FLT-PET), gives the potential opportunity to selectively avoid the portion of BM responsible for blood cells generation [15–18] Aim of the present planning comparison study is to test the hypothesis that the use of 18FDG- Page of 11 PET to identify pelvic active BM to be employed during automatic optimization process might enhance the chance to reduce the dose to the same structures compared to a planning process based on the wholebone delineation of pelvic bones This preliminary study aims at finding the most appropriate planning approach to be integrated within a prospective phase II trial in preparation at our Institute to decrease the hematologic toxicity profile in anal cancer patients undergoing CHTRT, employing dose-painted image-guided IMRT Methods Ten patients affected with locally advanced squamous cell carcinoma of the anal canal and/or margin were retrieved from our Institutional databased and employed for the present study In our center, 18FDG-PET-CT exam is prescribed to all anal cancer patients prior to treatment in order to complete the diagnostic and staging work-up These examinations were employed for our analysis Hence, it was not necessary to submit any patient to an extra diagnostic procedure for the present study Written informed consent was obtained from all patients, for 18FDG-PET-CT examination, radiotherapy treatment and clinical data utilization The Review Board of the Department of Oncology at the University of Turin approved the present study Overall patient and tumor characteristics are shown in Table Tumors were staged according to the 7th edition of the TNM classification (2010) Delineation of target volumes and organs at risk Patients had a virtual simulation procedure in supine position with both an indexed shaped knee rest and ankle support (CIVCO Medical Solutions, Kalona, IO, USA), without custom immobilization A CT scan was performed with mm slice thickness axial images acquired from the top of L1 vertebral body to the midfemural bones The gross tumor volume (GTV) comprised all primary and nodal macroscopic disease and was defined based on diagnostic MR and PET-CT images Primary and nodal GTVs were expanded isotropically with 20 mm and 10 mm respectively to generate the corresponding clinical target volumes (CTVs) and then modified to exclude osseous and muscular tissues The elective CTV encompassed the whole mesorectum and draining lymphatic regions, namely inguinal, external and internal iliac, obturator and perirectal nodes For locally advanced cases (cT4 and/or N2/N3), presacral nodes were also included within the CTV Lymphatic areas were contoured as a 10 mm isotropic expansion surrounding regional vessels and then modified to exclude bones and muscles Thereafter a 10 mm isotropic margin was added for the corresponding planning target volume (PTV) to account for organ motion and set up Franco et al BMC Cancer (2017) 17:710 errors Bladder, small and large bowel, external genitalia, femoral heads were defined as organs at risk (OARs) Radiotherapy dose prescription Dose prescriptions for target volumes were derived from Kachnic et al and adjusted according to clinical stage at presentation [6] Patients diagnosed with cT3-T4/N0-N3 disease were prescribed 54 Gy/30 fractions (1.8–2 Gy daily) to the anal gross tumor PTV, while gross nodal PTVs were prescribed 50.4 Gy/30 fr (1.68 Gy daily) if sized ≤3 cm or 54 Gy/30 fr (1.8 Gy daily) if >3 cm; elective nodal PTV was prescribed 45 Gy/30 fractions (1.5 Gy daily) [6] This is a frequently used fractionation to deliver IMRT treatments in this setting and it is a standard approach in our Institution [1–3, 5] This is the reason why it was chosen for the present study Chemotherapy All patients received concurrent CHT, consisting of 5fluorouracil (5-FU) (1000 mg/m2/day) given as continuous infusion along 96 h (days 1–5 and 29–33) associated with mitomycin C (MMC) (10 mg/m2, capped at maximum 20 mg single dose) given as bolus (days and 29) A total of concurrent cycles were administered Bone marrow delineation The external contour of pelvic bone marrow (PBM) was outlined on the planning CT using bone windows as first described by Mell et al [11] The PBM was delineated as a whole and then divided into subsites: a) the iliac BM (IBM), extending from the iliac crests to the upper border of femoral head; b) lower pelvis BM (LPBM), accounting for bilateral pube, ischia, acetabula and proximal femura, from the upper limit of the femoral heads to the lower limit of the ischial tuberosities and c) lumbosacral BM (LSBM), extending from the superior border of L5 somatic body [11] Active bone marrow delineation on FDG-PET All images derived from planning CT were exported on the Velocity platform (Varian Medical Systems, Palo Alto, CA) together with treatment volumes, OARs and dose references Given that FDG-PET-CT images were acquired separately, we performed a rigid co-registration between planning CT and PET-CT images Patients were set up in treatment position during the acquisition of FDG-PET-CT The 18FDG-PET standardized uptake values (SUVs) were calculated for PBM volumes, after correcting for body weight To standardize SUVs among all patients, we normalized BM and liver SUVs We defined as active bone marrow BM the volume having higher SUV values than the SUVmean for each patient, rather than the whole cohort, as proposed by Rose et al [19, 20] The areas identified with the method Page of 11 described above were outlined within PBM as a whole and named ACTPBM and within each of the subregions identified on planning CT (LSBM, IBM, LPBM) and named ACTLSBM, ACTIBM, ACTLPBM, respectively Inactive BM (1-ACTPBM) was identified as the difference between BM volumes as defined on planning CT and active BM The same procedure was done for all subregions to identify inactive BM within all of them The volumes were hence called 1-ACTLSBM, 1-ACTIBM, 1-ACTLPBM Planning process All treatment plans were generated using the Pinnacle3 ver 9.1 platform (Philips, Eindhoven, The Netherlands), including the Auto-planning (AP) module The AP engine is a progressive region of interest (ROI)-based optimization tool that creates all the required contours iteratively in order to optimize the dose distribution and takes PTV/OARs overlaps into account during the optimization process Moreover, AP is able to adjust the priority of clinical goals based on the probability to be achieved Besides clinical objectives and priorities, AP has a compromise setting to allow for sparing of serial organs such as the spinal cord over targets, and advanced settings to allow for setting global parameters such as priorities between targets and OARs, dose falloff, maximum dose and cold spot management Therefore the main input data required by AP to drive optimization are: target optimization goal, i.e prescription dose to the PTVs, engine type (biological or non biological), OARs optimization goals (max dose, max DVH or mean dose), priority (high, medium or low) and compromise (yes or no depending on the strength of the constraint) The standard OARs considered in the optimization process were: bladder (Dmax,Dmean,V35,V40,V50 as relative volumes), femural heads (Dmax,Dmean,V30,V40, as relative volumes), external genitalia (Dmax,Dmean,V20,V30,V40 as relative volumes), large and small bowel (Dmax,Dmean,V30,V45, as absolute volumes), iliac crests (V30,V40,V50 as relative volumes) and pelvic BM defined either as whole bone contour or using 18FDG-PET (lowest dose as possible) (Table 1) Four type of plans were created accounting for the various BM delineation approaches Each of the four trials was optimized considering BM as additional OAR (Fig 1): Plan A IBM (reference plan; accounting only for iliac crest as per RTOG 05–29 trial) Plan B IBM, LSBM, PBM and LPBM (accounting for all the pelvic BM as outlined by the outer surface of external osseous structures) Plan C ACTLSBM, ACTIBM, ACTLPBM, 1-ACTLSBM, 1-ACTIBM, 1-ACTLPBM (accounting for both the active BM subregions as defined by 18FDG-PET but also for the remaining parts of bony structures, to Franco et al BMC Cancer (2017) 17:710 Page of 11 Table Dose constraints to target volume and organs at risk employed during optimization PTV Bladder Large bowel Small bowel External genitalia Femural heads Iliac crests Parameter Goal D95% ≥95% Dmax ≤107% V30